Thanks, but I don't see how that link contradicts my objection. Can you point me to how they're tracking the workers that didn't end up being overqualified because they were excluded by the hiring managers?

From what I can tell, they tracked the jobs they took, but not the jobs they were denied for, so I don't see how they could've accounted for the selection bias.

That's because the selection bias you are asking them to account for is based on the psychic abilities of hiring managers. Even if it exists, it doesn't matter because there are still people hiring overqualified people and it only makes sense to talk about those in this context.

Either the effect exists or it doesn't, and you need to show that it doesn't if you want to claim that discriminating based on overqualification is not effective. And that study doesn't show that - it only shows that overqualified candidates that hiring managers thought wouldn't leave soon don't leave soon.